A. E. Verrill — Corals of the Genus Acropora. 235 



smaller, but several of the branches have them as large as is usual 

 in typical nohtlis j they vary from .} to 5'"'", the average being about 

 3 -5 to 4"". 



The structures of the caMienchyma and walls are as in vohilis. 



The lateral corallites are rather smaller than usual in that form, 

 and they are mostly distinctly compressed, and have the outer lip 

 less thickened, and more often incurved, so that many of the cor- 

 allites are slightly boat-shaped, and the calicles elliptical ; but they 

 are cut away on the upper side and have the short inner lip as in 

 Dobilis. 



Moreover, on some branches of the t^'^pe, the calicles are not more 

 compressed than often occurs on typical nohills. This character 

 varies in this species, as in many others, and may be due to more or 

 less crowding of the buds. 



This more compressed and more beaked form of the radial calicles 

 is, hoAvever, the only tangible character for separating this form, even 

 as a variet3% 



The seeund condition, due to the partial suppression of the coral- 

 lites on the under sides of some branches, is an accident of growth 

 that may occur in any species. The smaller size of the branches is 

 not even of varietal importance. 



Brook not only considers secnnda a good species, but he puts it in 

 a different group,* far removed from nobilis, though he refers to my 

 uniting these forms in 1864. Probably Brook had a different species, 

 from Australia,! which he described under the name of secunda ,' 

 but he also quotes Dana's description. 



My conclusion in regard to this point, in 1864, was based on a 

 direct cotnparison. of DanrCs types of both forms, with a fine series 

 of nobilis in the Mus. of Comp. Zoology. Although I have studied 

 larger collections since then, I have seen no reason to change my 

 opinion. 



■" Brook puts secunda in his 1st subgenus, Euniadrepora, on p. 30 ; nobilis in 

 his 6th subgenus, Tylopora, on p. 135. This is mainly an account of the slight 

 diflference in the size of the axial corallites. 



I Brook's Australian specimen is said to have the radial calicles rather distant, 

 miich compressed, thin-walled, tubo-nariform or dimidiate, 3-4.5""™ long, 1.2- 

 1.6""" broad ; the walls striato-reticulate, not echinulate, iiuless at base. Most 

 of these characters do not apply to Dana's secunda, especially the thin-walled, 

 elongated, nariform corallites, nor would Dana's species go in the subgenus 

 Eumadrepora, as defined. 



Hence I believe it a distinct species with a superficial resemblance to secunda 

 and would propose to call it secnndcUa, sp. liov. The types are from Port Deni- 

 son and Bandin Is., Australia (coll. Kent). 



